Does rural electrification cause economic development?

The last unelectrified Indian village is now connected to the grid. Given the large investments required for these infrastructure projects, value-for-money is an important consideration. Examining the impact of the Rajiv Gandhi Grameen Vidyutikaran Yojana, which expanded electricity access in 400,000 villages during 2005-2011, this article shows that the economic benefits of the intervention likely do not outweigh the costs below a certain population threshold.

Nearly one billion people, most of whom reside in rural communities in South Asia and sub-Saharan Africa, still lack modern electricity access. Expanding energy infrastructure to cover everyone on the planet by 2030 is one of the United Nations’ 17 global Sustainable Development Goals, and would cost an estimated US$49 billion per year (International Energy Agency, 2019).

This push comes in the context of mixed evidence on the causal effects of rural electrification on economic development. Early research found large positive impacts of electrification on development outcomes (Dinkelman 2011, Rud 2012, Lipscomb et al. 2013). At the same time, recent experimental evidence shows that continuing to expand grid power access would yield at best small gains in economic welfare (Lee et al. 2020, Burgess et al. 2023). These discrepancies may reflect differences in scale: while estimates of large benefits have tended to come from electrifying large communities, districts, or even states, estimates of small benefits have come from village-level electrification – a scale that is more representative of today’s electrification efforts.

In our research (Burlig and Preonas 2024), we study the world’s largest-ever rural electrification push: India’s Rajiv Gandhi Grameen Vidyutikaran Yojana (RGGVY) programme, which expanded electricity access in over 400,000 rural villages between 2005 and 2011. We leverage two natural experiments created by the design of the RGGVY programme: (i) A population-based eligibility cutoff, facilitating a regression discontinuity design (RDD)¹, and (ii) A staggered rollout across districts, facilitating a difference-in-differences (DID) design². Using this quasi-random policy variation, we estimate credible causal effects of rural electrification among the world’s largest under-electrified population.

The RGGVY Programme

Rajiv Gandhi Grameen Vidyutikaran Yojana, or the “Rajiv Gandhi Rural Electrification Programme”, was launched in 2005. At that time, over 125,000 rural villages still lacked power access, and 57% of all rural households in India lacked grid connections. RGGVY sought to connect over 100,000 previously unelectrified villages, and to expand electricity access in 300,000 ‘under-electrified’ villages. The programme constructed new transmission lines, distribution lines, and transformers, and was charged with providing free grid connections to below-poverty-line households. By 2011, RGGVY had connected 17.5 million households to the grid, or one in five previously unelectrified rural households in India (Sreekumar and Dixit 2011).

RGGVY funds were released to districts in two waves, on a first-come-first-served basis. This staggered rollout governed the timing of electrification, and our DID design compares outcomes for districts in the first and second RGGVY funding waves. Our administrative outcome data come from 2011, which was 3-5 years after first-wave districts received treatment under RGGVY, but crucially before the completion of RGGVY projects in second-wave districts.

RGGVY included a population-based eligibility threshold: villages below 300 people were ineligible for grid connections under the scheme, which enables us to compare villages with populations just above and below this threshold using the RDD methodology. Importantly, village populations were recorded prior to the RGGVY programme, precluding villages from manipulating their population statistics to qualify for the programme (which we nevertheless test for and do not detect).

Effect of the programme on electricity access

Using data from the 2011 Census which capture the extensive margin of electrification, we estimate that RGGVY caused a 10% increase in commercial power access among eligible villages, aligned with the programme’s goal of supporting microenterprises. Eligible villages also saw a 14% increase in the hours of commercial power availability.

We use annual satellite images of night-time brightness to show the change in electricity access over time. This reveals brightness effects that increase in each year post electrification (Figure 1). By using ‘groundtruthed’ estimates (that is, corresponding to the actual location on Earth) based on previous research using remote sensing data (for example, Machemedze et al. 2017), our estimates imply a 10 percentage point increase in the share of rural households with electric lighting.

Figure 1. Village-level RDD estimates in night-time brightness, by year

Notes: (i) This figure plots RDD point estimates from 11 separate RDD regressions, where the outcome variables are village-level, night-time brightness in each year. (ii) The vertical lines indicate 95% confidence intervals. A confidence interval is a way of expressing uncertainty about estimated effects – specifically, it means that if the study was repeated over and over with new samples, 95% of the time the calculated confidence interval would contain the true effect. (iii) We see no difference in night-time brightness between eligible and ineligible villages prior to the RGGVY programme’s start. We begin to detect statistically significant impacts in 2008, which was less than two years after first-wave villages received treatment under RGGVY.

Using household survey data that is representative at the district level, we find direct evidence that RGGVY increased household electrification. We estimate the following effects: RGGVY caused a 9% increase in household grid connections, a 13% increase in households’ monthly electricity consumption, an 8% increase in electric lighting, and an 11% increase in electric fan ownership. These estimates are all statistically significant and highly robust. Together, they confirm that the RGGVY programme created economically meaningful gains in rural electrification – while also falling short of the ‘full electrification’ ideal.

Economic benefits of electrification relative to village size

Next, we estimate RGGVY’s economic impacts. We find that electrification had no effects on a wide range of economic outcomes, including village demographics, employment, and education. Our ‘precise null’ estimates, using both of our research designs, mean that we can statistically reject that electrification caused even small increases in household consumption expenditures, our preferred economic outcome.

We then scale-up these RGGVY-specific results to consider the welfare impacts of a hypothetical full electrification programme that connected all households to the grid. This exercise reveals that full electrification would have no economic benefits for smaller villages (of between 300 and 1,000 people).

However, we do find evidence of potentially large economic gains from full electrification in larger communities: our analysis for 2,000-person villages suggests that full electrification could more than double expenditures. This aligns with our estimates of RGGVY’s impact on microenterprises: we find null effects for small villages, but statistically significant 10% increases in the number of firms in large villages. This suggests that only large villages were able to reap the benefits of expanded electricity access by shifting production into firms.

Welfare analysis of rural electrification

Finally, we use these estimates to quantify the internal rate of return (IRR) from full electrification, net of programme costs (Figure 2). For a 300-person village, our estimates imply a 0% IRR, meaning that even without any time discounting, the 20-year flow of household surplus from electricity consumption would not exceed upfront electrification costs. For a 1,000-person village, we recover a 13% IRR, which barely exceeds the standard 10-12% benchmark commonly used to assess cost-effectiveness of development programmes (Asian Development Bank, 2013). For a 2,000-person village, we recover a 33% IRR, which far exceeds this benchmark and suggests substantial welfare benefits from full electrification.

Figure 2. Implied internal rate of return from full electrification, by village size

Note: This figure graphs the discount rate required for the present value of future household benefits from electricity consumption to outweigh the upfront costs of electrification.

Three factors explain why these welfare effects hinge on village size: in larger villages, (i) there are greater per capita benefits from electrification, (ii) these benefits accrue to more people, and (iii) the fixed costs of electricity infrastructure are spread across more grid connections. Our findings help to resolve a fundamental puzzle in the literature on the economics of rural electrification, since community population size appears to explain much of the disagreement between previous empirical estimates on the economic impacts of expanding grid power.

Policy implications

Electrification pushes are continuing worldwide, with India recently having connected its last unelectrified village to the grid. Given the scope of spending on these large infrastructure projects, governments should target electrification where the benefits are likely to be highest. Our results suggest that population size can serve as a key selection criterion for ongoing electrification efforts. For smaller villages that remain under-electrified, alternative technologies such as solar microgrids may offer a more cost-effective means of expanding energy access than grid power.

This article first appeared on VoxDev.

I4I is now on Substack. Please click here (@Ideas for India) to subscribe to our channel for quick updates on our content

Notes:

Regression discontinuity is a quasi-experimental impact evaluation method that isolates the causal effect of a treatment (intervention) by comparing outcomes just above and below an eligibility threshold – in this case, the population cutoff – that determines the assignment of treatment.
Difference-in-differences is a technique used to compare the evolution of outcomes over time between two similar groups, wherein one group gains access to an intervention in the first instance while the other group does not.